vnet/vnet/devices/dpdk/node.c

/*
 * Copyright (c) 2015 Cisco and/or its affiliates.
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at:
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */
#include <vnet/vnet.h>
#include <vppinfra/vec.h>
#include <vppinfra/error.h>
#include <vppinfra/format.h>
#include <vppinfra/xxhash.h>

#include <vnet/ethernet/ethernet.h>
#include <vnet/devices/dpdk/dpdk.h>
#include <vnet/classify/vnet_classify.h>
#include <vnet/mpls/packet.h>
#include <vnet/handoff.h>
#include <vnet/devices/devices.h>
#include <vnet/feature/feature.h>

#include "dpdk_priv.h"

#ifndef MAX
#define MAX(a,b) ((a) < (b) ? (b) : (a))
#endif

#ifndef MIN
#define MIN(a,b) ((a) < (b) ? (a) : (b))
#endif

/*
 * At least in certain versions of ESXi, vmware e1000's don't honor the
 * "strip rx CRC" bit. Set this flag to work around that bug FOR UNIT TEST ONLY.
 *
 * If wireshark complains like so:
 *
 * "Frame check sequence: 0x00000000 [incorrect, should be <hex-num>]"
 * and you're using ESXi emulated e1000's, set this flag FOR UNIT TEST ONLY.
 *
 * Note: do NOT check in this file with this workaround enabled! You'll lose
 * actual data from e.g. 10xGE interfaces. The extra 4 bytes annoy
 * wireshark, but they're harmless...
 */
#define VMWARE_LENGTH_BUG_WORKAROUND 0

static char *dpdk_error_strings[] = {
#define _(n,s) s,
  foreach_dpdk_error
#undef _
};

always_inline int
dpdk_mbuf_is_ip4 (struct rte_mbuf *mb)
{
  return RTE_ETH_IS_IPV4_HDR (mb->packet_type) != 0;
}

always_inline int
dpdk_mbuf_is_ip6 (struct rte_mbuf *mb)
{
  return RTE_ETH_IS_IPV6_HDR (mb->packet_type) != 0;
}

always_inline int
vlib_buffer_is_mpls (vlib_buffer_t * b)
{
  ethernet_header_t *h = (ethernet_header_t *) b->data;
  return (h->type == clib_host_to_net_u16 (ETHERNET_TYPE_MPLS_UNICAST));
}

always_inline void
dpdk_rx_next_and_error_from_mb_flags_x1 (dpdk_device_t * xd,
					 struct rte_mbuf *mb,
					 vlib_buffer_t * b0, u32 * next0,
					 u8 * error0)
{
  u8 n0;
  uint16_t mb_flags = mb->ol_flags;

  if (PREDICT_FALSE (mb_flags & (
#ifdef RTE_LIBRTE_MBUF_EXT_RX_OLFLAGS
				  PKT_EXT_RX_PKT_ERROR | PKT_EXT_RX_BAD_FCS |
#endif /* RTE_LIBRTE_MBUF_EXT_RX_OLFLAGS */
				  PKT_RX_IP_CKSUM_BAD | PKT_RX_L4_CKSUM_BAD)))
    {
      /* some error was flagged. determine the drop reason */
      n0 = VNET_DEVICE_INPUT_NEXT_DROP;
      *error0 =
#ifdef RTE_LIBRTE_MBUF_EXT_RX_OLFLAGS
	(mb_flags & PKT_EXT_RX_PKT_ERROR) ? DPDK_ERROR_RX_PACKET_ERROR :
	(mb_flags & PKT_EXT_RX_BAD_FCS) ? DPDK_ERROR_RX_BAD_FCS :
#endif /* RTE_LIBRTE_MBUF_EXT_RX_OLFLAGS */
	(mb_flags & PKT_RX_IP_CKSUM_BAD) ? DPDK_ERROR_IP_CHECKSUM_ERROR :
	(mb_flags & PKT_RX_L4_CKSUM_BAD) ? DPDK_ERROR_L4_CHECKSUM_ERROR :
	DPDK_ERROR_NONE;
    }
  else
    {
      *error0 = DPDK_ERROR_NONE;
      if (PREDICT_FALSE (xd->per_interface_next_index != ~0))
	{
	  n0 = xd->per_interface_next_index;
	  b0->flags |= BUFFER_HANDOFF_NEXT_VALID;
	  if (PREDICT_TRUE (dpdk_mbuf_is_ip4 (mb)))
	    vnet_buffer (b0)->handoff.next_index =
	      HANDOFF_DISPATCH_NEXT_IP4_INPUT;
	  else if (PREDICT_TRUE (dpdk_mbuf_is_ip6 (mb)))
	    vnet_buffer (b0)->handoff.next_index =
	      HANDOFF_DISPATCH_NEXT_IP6_INPUT;
	  else if (PREDICT_TRUE (vlib_buffer_is_mpls (b0)))
	    vnet_buffer (b0)->handoff.next_index =
	      HANDOFF_DISPATCH_NEXT_MPLS_INPUT;
	  else
	    vnet_buffer (b0)->handoff.next_index =
	      HANDOFF_DISPATCH_NEXT_ETHERNET_INPUT;
	}
      else
	if (PREDICT_FALSE ((xd->flags & DPDK_DEVICE_FLAG_HAVE_SUBIF) ||
			   (mb_flags & PKT_RX_VLAN_PKT)))
	n0 = VNET_DEVICE_INPUT_NEXT_ETHERNET_INPUT;
      else
	{
	  if (PREDICT_TRUE (dpdk_mbuf_is_ip4 (mb)))
	    n0 = VNET_DEVICE_INPUT_NEXT_IP4_INPUT;
	  else if (PREDICT_TRUE (dpdk_mbuf_is_ip6 (mb)))
	    n0 = VNET_DEVICE_INPUT_NEXT_IP6_INPUT;
	  else if (PREDICT_TRUE (vlib_buffer_is_mpls (b0)))
	    n0 = VNET_DEVICE_INPUT_NEXT_MPLS_INPUT;
	  else
	    n0 = VNET_DEVICE_INPUT_NEXT_ETHERNET_INPUT;
	}
    }
  *next0 = n0;
}

void
dpdk_rx_trace (dpdk_main_t * dm,
	       vlib_node_runtime_t * node,
	       dpdk_device_t * xd,
	       u16 queue_id, u32 * buffers, uword n_buffers)
{
  vlib_main_t *vm = vlib_get_main ();
  u32 *b, n_left;
  u32 next0;

  n_left = n_buffers;
  b = buffers;

  while (n_left >= 1)
    {
      u32 bi0;
      vlib_buffer_t *b0;
      dpdk_rx_dma_trace_t *t0;
      struct rte_mbuf *mb;
      u8 error0;

      bi0 = b[0];
      n_left -= 1;

      b0 = vlib_get_buffer (vm, bi0);
      mb = rte_mbuf_from_vlib_buffer (b0);
      dpdk_rx_next_and_error_from_mb_flags_x1 (xd, mb, b0, &next0, &error0);
      vlib_trace_buffer (vm, node, next0, b0, /* follow_chain */ 0);
      t0 = vlib_add_trace (vm, node, b0, sizeof (t0[0]));
      t0->queue_index = queue_id;
      t0->device_index = xd->device_index;
      t0->buffer_index = bi0;

      clib_memcpy (&t0->mb, mb, sizeof (t0->mb));
      clib_memcpy (&t0->buffer, b0, sizeof (b0[0]) - sizeof (b0->pre_data));
      clib_memcpy (t0->buffer.pre_data, b0->data,
		   sizeof (t0->buffer.pre_data));
      clib_memcpy (&t0->data, mb->buf_addr + mb->data_off, sizeof (t0->data));

#ifdef RTE_LIBRTE_MBUF_EXT_RX_OLFLAGS
      /*
       * Clear overloaded TX offload flags when a DPDK driver
       * is using them for RX flags (e.g. Cisco VIC Ethernet driver)
       */
      mb->ol_flags &= PKT_EXT_RX_CLR_TX_FLAGS_MASK;
#endif /* RTE_LIBRTE_MBUF_EXT_RX_OLFLAGS */

      b += 1;
    }
}

static inline u32
dpdk_rx_burst (dpdk_main_t * dm, dpdk_device_t * xd, u16 queue_id)
{
  u32 n_buffers;
  u32 n_left;
  u32 n_this_chunk;

  n_left = VLIB_FRAME_SIZE;
  n_buffers = 0;

  if (PREDICT_TRUE (xd->flags & DPDK_DEVICE_FLAG_PMD))
    {
      while (n_left)
	{
	  n_this_chunk = rte_eth_rx_burst (xd->device_index, queue_id,
					   xd->rx_vectors[queue_id] +
					   n_buffers, n_left);
	  n_buffers += n_this_chunk;
	  n_left -= n_this_chunk;

	  /* Empirically, DPDK r1.8 produces vectors w/ 32 or fewer elts */
	  if (n_this_chunk < 32)
	    break;
	}
    }
  else
    {
      ASSERT (0);
    }

  return n_buffers;
}

/*
 * This function is used when there are no worker threads.
 * The main thread performs IO and forwards the packets.
 */
static inline u32
dpdk_device_input (dpdk_main_t * dm,
		   dpdk_device_t * xd,
		   vlib_node_runtime_t * node,
		   u32 cpu_index, u16 queue_id)
{
  u32 n_buffers;
  u32 next_index = VNET_DEVICE_INPUT_NEXT_ETHERNET_INPUT;
  u32 n_left_to_next, *to_next;
  u32 mb_index;
  vlib_main_t *vm = vlib_get_main ();
  uword n_rx_bytes = 0;
  u32 n_trace, trace_cnt __attribute__ ((unused));
  vlib_buffer_free_list_t *fl;
  u32 buffer_flags_template;

  if ((xd->flags & DPDK_DEVICE_FLAG_ADMIN_UP) == 0)
    return 0;

  n_buffers = dpdk_rx_burst (dm, xd, queue_id);

  if (n_buffers == 0)
    {
      return 0;
    }

  buffer_flags_template = dm->buffer_flags_template;

  vec_reset_length (xd->d_trace_buffers);
  trace_cnt = n_trace = vlib_get_trace_count (vm, node);

  fl = vlib_buffer_get_free_list (vm, VLIB_BUFFER_DEFAULT_FREE_LIST_INDEX);

  mb_index = 0;

  while (n_buffers > 0)
    {
      u32 bi0, next0;
      u8 error0;
      u32 l3_offset0;
      vlib_buffer_t *b0, *b_seg, *b_chain = 0;

      vlib_get_next_frame (vm, node, next_index, to_next, n_left_to_next);

      while (n_buffers > 0 && n_left_to_next > 0)
	{
	  u8 nb_seg = 1;
	  struct rte_mbuf *mb = xd->rx_vectors[queue_id][mb_index];
	  struct rte_mbuf *mb_seg = mb->next;

	  if (PREDICT_TRUE (n_buffers > 2))
	    {
	      struct rte_mbuf *pfmb = xd->rx_vectors[queue_id][mb_index + 2];
	      vlib_buffer_t *bp = vlib_buffer_from_rte_mbuf (pfmb);
	      CLIB_PREFETCH (pfmb, CLIB_CACHE_LINE_BYTES, STORE);
	      CLIB_PREFETCH (bp, CLIB_CACHE_LINE_BYTES, STORE);
	    }

	  ASSERT (mb);

	  b0 = vlib_buffer_from_rte_mbuf (mb);

	  /* Prefetch one next segment if it exists. */
	  if (PREDICT_FALSE (mb->nb_segs > 1))
	    {
	      struct rte_mbuf *pfmb = mb->next;
	      vlib_buffer_t *bp = vlib_buffer_from_rte_mbuf (pfmb);
	      CLIB_PREFETCH (pfmb, CLIB_CACHE_LINE_BYTES, LOAD);
	      CLIB_PREFETCH (bp, CLIB_CACHE_LINE_BYTES, STORE);
	      b_chain = b0;
	    }

	  vlib_buffer_init_for_free_list (b0, fl);

	  bi0 = vlib_get_buffer_index (vm, b0);

	  to_next[0] = bi0;
	  to_next++;
	  n_left_to_next--;

	  dpdk_rx_next_and_error_from_mb_flags_x1 (xd, mb, b0,
						   &next0, &error0);
#ifdef RTE_LIBRTE_MBUF_EXT_RX_OLFLAGS
	  /*
	   * Clear overloaded TX offload flags when a DPDK driver
	   * is using them for RX flags (e.g. Cisco VIC Ethernet driver)
	   */

	  if (PREDICT_TRUE (trace_cnt == 0))
	    mb->ol_flags &= PKT_EXT_RX_CLR_TX_FLAGS_MASK;
	  else
	    trace_cnt--;
#endif /* RTE_LIBRTE_MBUF_EXT_RX_OLFLAGS */

	  b0->error = node->errors[error0];

	  l3_offset0 = ((next0 == VNET_DEVICE_INPUT_NEXT_IP4_INPUT ||
			 next0 == VNET_DEVICE_INPUT_NEXT_IP6_INPUT ||
			 next0 == VNET_DEVICE_INPUT_NEXT_MPLS_INPUT) ?
			sizeof (ethernet_header_t) : 0);

	  b0->current_data = l3_offset0;
	  /* Some drivers like fm10k receive frames with
	     mb->data_off > RTE_PKTMBUF_HEADROOM */
	  b0->current_data += mb->data_off - RTE_PKTMBUF_HEADROOM;
	  b0->current_length = mb->data_len - l3_offset0;

	  b0->flags = buffer_flags_template;

	  if (VMWARE_LENGTH_BUG_WORKAROUND)
	    b0->current_length -= 4;

	  vnet_buffer (b0)->sw_if_index[VLIB_RX] = xd->vlib_sw_if_index;
	  vnet_buffer (b0)->sw_if_index[VLIB_TX] = (u32) ~ 0;
	  n_rx_bytes += mb->pkt_len;

	  /* Process subsequent segments of multi-segment packets */
	  while ((mb->nb_segs > 1) && (nb_seg < mb->nb_segs))
	    {
	      ASSERT (mb_seg != 0);

	      b_seg = vlib_buffer_from_rte_mbuf (mb_seg);
	      vlib_buffer_init_for_free_list (b_seg, fl);

	      ASSERT ((b_seg->flags & VLIB_BUFFER_NEXT_PRESENT) == 0);
	      ASSERT (b_seg->current_data == 0);

	      /*
	       * The driver (e.g. virtio) may not put the packet data at the start
	       * of the segment, so don't assume b_seg->current_data == 0 is correct.
	       */
	      b_seg->current_data =
		(mb_seg->buf_addr + mb_seg->data_off) - (void *) b_seg->data;

	      b_seg->current_length = mb_seg->data_len;
	      b0->total_length_not_including_first_buffer += mb_seg->data_len;

	      b_chain->flags |= VLIB_BUFFER_NEXT_PRESENT;
	      b_chain->next_buffer = vlib_get_buffer_index (vm, b_seg);

	      b_chain = b_seg;
	      mb_seg = mb_seg->next;
	      nb_seg++;
	    }

	  /*
	   * Turn this on if you run into
	   * "bad monkey" contexts, and you want to know exactly
	   * which nodes they've visited... See main.c...
	   */
	  VLIB_BUFFER_TRACE_TRAJECTORY_INIT (b0);

          /* Do we have any driver RX features configured on the interface? */
	  vnet_feature_start_device_input_x1 (xd->vlib_sw_if_index, &next0, b0, l3_offset0);

	  vlib_validate_buffer_enqueue_x1 (vm, node, next_index,
					   to_next, n_left_to_next,
					   bi0, next0);
	  if (PREDICT_FALSE (n_trace > mb_index))
	    vec_add1 (xd->d_trace_buffers, bi0);
	  n_buffers--;
	  mb_index++;
	}
      vlib_put_next_frame (vm, node, next_index, n_left_to_next);
    }

  if (PREDICT_FALSE (vec_len (xd->d_trace_buffers) > 0))
    {
      dpdk_rx_trace (dm, node, xd, queue_id, xd->d_trace_buffers,
		     vec_len (xd->d_trace_buffers));
      vlib_set_trace_count (vm, node,
			    n_trace - vec_len (xd->d_trace_buffers));
    }

  vlib_increment_combined_counter
    (vnet_get_main ()->interface_main.combined_sw_if_counters
     + VNET_INTERFACE_COUNTER_RX,
     cpu_index, xd->vlib_sw_if_index, mb_index, n_rx_bytes);

  dpdk_worker_t *dw = vec_elt_at_index (dm->workers, cpu_index);
  dw->aggregate_rx_packets += mb_index;

  return mb_index;
}

static inline void
poll_rate_limit (dpdk_main_t * dm)
{
  /* Limit the poll rate by sleeping for N msec between polls */
  if (PREDICT_FALSE (dm->poll_sleep != 0))
    {
      struct timespec ts, tsrem;

      ts.tv_sec = 0;
      ts.tv_nsec = 1000 * 1000 * dm->poll_sleep;	/* 1ms */

      while (nanosleep (&ts, &tsrem) < 0)
	{
	  ts = tsrem;
	}
    }
}

/** \brief Main DPDK input node
    @node dpdk-input

    This is the main DPDK input node: across each assigned interface,
    call rte_eth_rx_burst(...) or similar to obtain a vector of
    packets to process. Handle early packet discard. Derive @c
    vlib_buffer_t metadata from <code>struct rte_mbuf</code> metadata,
    Depending on the resulting metadata: adjust <code>b->current_data,
    b->current_length </code> and dispatch directly to
    ip4-input-no-checksum, or ip6-input. Trace the packet if required.

    @param vm   vlib_main_t corresponding to the current thread
    @param node vlib_node_runtime_t
    @param f    vlib_frame_t input-node, not used.

    @par Graph mechanics: buffer metadata, next index usage

    @em Uses:
    - <code>struct rte_mbuf mb->ol_flags</code>
        - PKT_EXT_RX_PKT_ERROR, PKT_EXT_RX_BAD_FCS
        PKT_RX_IP_CKSUM_BAD, PKT_RX_L4_CKSUM_BAD
    - <code> RTE_ETH_IS_xxx_HDR(mb->packet_type) </code>
        - packet classification result

    @em Sets:
    - <code>b->error</code> if the packet is to be dropped immediately
    - <code>b->current_data, b->current_length</code>
        - adjusted as needed to skip the L2 header in  direct-dispatch cases
    - <code>vnet_buffer(b)->sw_if_index[VLIB_RX]</code>
        - rx interface sw_if_index
    - <code>vnet_buffer(b)->sw_if_index[VLIB_TX] = ~0</code>
        - required by ipX-lookup
    - <code>b->flags</code>
        - to indicate multi-segment pkts (VLIB_BUFFER_NEXT_PRESENT), etc.

    <em>Next Nodes:</em>
    - Static arcs to: error-drop, ethernet-input,
      ip4-input-no-checksum, ip6-input, mpls-input
    - per-interface redirection, controlled by
      <code>xd->per_interface_next_index</code>
*/

static uword
dpdk_input (vlib_main_t * vm, vlib_node_runtime_t * node, vlib_frame_t * f)
{
  dpdk_main_t *dm = &dpdk_main;
  dpdk_device_t *xd;
  uword n_rx_packets = 0;
  dpdk_device_and_queue_t *dq;
  u32 cpu_index = os_get_cpu_number ();

  /*
   * Poll all devices on this cpu for input/interrupts.
   */
  /* *INDENT-OFF* */
  vec_foreach (dq, dm->devices_by_cpu[cpu_index])
    {
      xd = vec_elt_at_index(dm->devices, dq->device);
      ASSERT(dq->queue_id == 0);
      n_rx_packets += dpdk_device_input (dm, xd, node, cpu_index, 0);
    }
  /* *INDENT-ON* */

  poll_rate_limit (dm);

  return n_rx_packets;
}

uword
dpdk_input_rss (vlib_main_t * vm,
		vlib_node_runtime_t * node, vlib_frame_t * f)
{
  dpdk_main_t *dm = &dpdk_main;
  dpdk_device_t *xd;
  uword n_rx_packets = 0;
  dpdk_device_and_queue_t *dq;
  u32 cpu_index = os_get_cpu_number ();

  /*
   * Poll all devices on this cpu for input/interrupts.
   */
  /* *INDENT-OFF* */
  vec_foreach (dq, dm->devices_by_cpu[cpu_index])
    {
      xd = vec_elt_at_index(dm->devices, dq->device);
      n_rx_packets += dpdk_device_input (dm, xd, node, cpu_index, dq->queue_id);
    }
  /* *INDENT-ON* */

  poll_rate_limit (dm);

  return n_rx_packets;
}

/* *INDENT-OFF* */
VLIB_REGISTER_NODE (dpdk_input_node) = {
  .function = dpdk_input,
  .type = VLIB_NODE_TYPE_INPUT,
  .name = "dpdk-input",
  .sibling_of = "device-input",

  /* Will be enabled if/when hardware is detected. */
  .state = VLIB_NODE_STATE_DISABLED,

  .format_buffer = format_ethernet_header_with_length,
  .format_trace = format_dpdk_rx_dma_trace,

  .n_errors = DPDK_N_ERROR,
  .error_strings = dpdk_error_strings,
};


/* handle dpdk_input_rss alternative function */
VLIB_NODE_FUNCTION_MULTIARCH_CLONE(dpdk_input)
VLIB_NODE_FUNCTION_MULTIARCH_CLONE(dpdk_input_rss)

/* this macro defines dpdk_input_rss_multiarch_select() */
CLIB_MULTIARCH_SELECT_FN(dpdk_input);
CLIB_MULTIARCH_SELECT_FN(dpdk_input_rss);